I am having my first steps in EF 4.1. Because I was using NHibenate, the code first approach seems to me as the best one. I have problem with good mapping of one-to-many (or many-to-one) realtionship. Let's say I have 2 entities:
class ClientModel
{
int ClientID;
string Name;
virtual IList<OrderModel> Orders;
}
class OrderModel
{
int OrderID;
string Details;
virtual ClienModel Client;
}
When I leave it like that, there is an error while generating database - keys in tables are missing. I figured out I can fix it by changing names of the keys to ID (but it's not OK with my naming convention) or by adding [Key] annotation. Even if I add this annotation, still the names of tables are wrong - just like classes names but with 's'.
So I tried to use fluent API - I made mappings. But if I set mappings just like here:
class ClientMapping
{
ClientMapping()
{
this.HasKey(e => e.ClientID).Property(e => e.ID).HasColumnName("ClientID");
this.Property(e => e.Name).HasColumnName("Name");
this.HasMany(e => e.Orders).WithOptional().Map(p => p.MapKey("OrderID")).WillCascadeOnDelete();
this.ToTable("Clients");
}
}
class OrderMapping
{
OrderMapping()
{
this.HasKey(e => e.OrderID).Property(e => e.OrderID).HasColumnName("OrderID");
this.Property(e => e.Details).HasColumnName("Details");
this.HasRequired(e => e.Client).WithMany().Map(p=>p.MapKey("Client")).WillCascadeOnDelete(false);
this.ToTable("Orders");
}
}
the relation betweene tables in database is doubled.
What is the proper way to do one-to-many relationship using code-first approach? Am I thinking in a good direction or is it a wrong approach?
EDIT
OK, I have done it in the way #Eranga showed, but there is still a problem. When I'm getting Client from database, its Orders property is null (but in database it has some Orders with Order.ClientID == Client.ClientID).
You need to map both properties participating in the relationship. You need to add ClientID column to Orders table.
class ClientMapping
{
ClientMapping()
{
this.HasKey(e => e.ClientID).Property(e => e.ID).HasColumnName("ClientID");
this.Property(e => e.Name).HasColumnName("Name");
this.HasMany(e => e.Orders).WithRequired(o => o.Client)
.Map(p => p.MapKey("ClientID")).WillCascadeOnDelete();
this.ToTable("Clients");
}
}
class OrderMapping
{
OrderMapping()
{
this.HasKey(e => e.OrderID).Property(e => e.OrderID).HasColumnName("OrderID");
this.Property(e => e.Details).HasColumnName("Details");
this.ToTable("Orders");
}
}
Configuring the relationship from one entity is sufficient.
This may help (it helped me, when i couldn't figure out how this works):
If you would have the classes like this:
class ClientModel
{
int ClientId;
string Name;
}
class OrderModel
{
int OrderId;
string Details;
int ClientId;
}
Then this would represent 2 tables in your database which "wouldn't" be connected with each other via a foreign key (they would be connected via the ClientId in the OrderModel) and you could get data like "GetAllOrdersWithSomeClientId" and "GetTheClientNameForSomeClientId" from the database. BUT problems would arise when you would delete a Client from the database. Because then there would still be some Orders which would contain a ClientId which doesn't exist in the Client table anymore and that would lead to anomalies in your database.
The virtual List<OrderModel> Orders; (in the ClientModel) and virtual ClienModel Client; (in the OrderModel) are needed to create the relation aka. the foreign key between the tables ClientModel and OrderModel.
There is one thing about which i'm still not sure about. Which is the int ClientId; in the OrderModel. I guess that it has to have the same name as the ClientId in the ClientModel so that the entity framework knows which 2 attributes the foreign key has to connect. Would be nice if someone could explain this in detail.
Also, put this into your DbContext constructor if something souldn't work:
this.Configuration.ProxyCreationEnabled = false;
this.Configuration.LazyLoadingEnabled = false;
Related
Using EF Core 5.0, I have a PK-less entity (from a SQL view) OrderInfo, which has a column OrderDetailId. I also have an entity DiscountOrder which a PK from the columns OrderDetailId and DiscountId.
I would like to create a navigation property from Order to DiscountOrders. Such as:
public class OrderInfo
{
public int OrderDetailId { get; set; }
public virtual ICollection<DiscountOrder> DiscountOrders { get; set; }
}
public class DiscountOrder
{
public int DiscountId { get; set; }
public int OrderDetailId { get; set; }
}
// For reference, this entity also exists
public class Discount
{
public int DiscountId { get; set; }
}
Obviously, there are no FKs to make use of, but I should be able to create a navigation property anyway.
I think I should be able to do this:
modelBuilder.Entity<OrderInfo>(e =>
{
e.HasNoKey();
e.HasMany(x => x.DiscountOrders)
.WithOne()
.HasPrincipalKey(o => o.OrderDetailId)
.HasForeignKey(pb => pb.OrderDetailId)
.IsRequired(false);
});
But a query on DbSet<OrderInfo> results in a NullReferenceException with the breakpoint landing on the HasMany() line. That said, I don't do anything with the DiscountOrders property, so the exception seems like it would have to be configuration related.
I've looked at answers to similar questions, but most answers use HasOne().WithMany() where as I'd like to keep this definition on OrderInfo since we don't really care about the other direction. How can I correctly set up this mapping?
Keyless entities (entity without key) cannot be principal of a relationship, because there is no key to be referenced by the FK property of the dependent.
Note that by EF Core terminology key is primary key. There are also alternate (unique) keys, but EF Core does not enable them for keyless types.
So basically HasNoKey() disables alternate keys and relationships to that entity. Just the exception is unhandled, hence not user friendly. For instance, if you try to predefine the alternate key referenced by .HasPrincipalKey(o => o.OrderDetailId) in advance
e.HasNoKey();
e.HasAlternateKey(o => o.OrderDetailId);
you'll get much better exception message at the second line
The key {'OrderDetailId'} cannot be added to keyless type 'OrderInfo'.
Shortly, e.HasNoKey(); and `.HasPrincipalKey(o => o.OrderDetailId); are mutually exclusive.
The only way to make it work is to define PK for OrderInfo even though it does not exist in database. In fact if OrderDetailId was supposed to be alternate key, in other words, is unique in the returned set, then you can safely map it as PK
//e.HasNoKey();
e.HasKey(o => o.OrderDetailId);
If it is not unique, then nothing can be done - you cannot map and use navigation property, and will be forced to use manual joins in L2E queries.
Update: EF Core also blocks changing "keyless"-ness once it's been set via fluent API (which has the highest configuration priority). So if you can't remove HasNoKey() fluent call because of it being generated by reverse engineering, you have to resort to metadata API to make it again "normal" entity by setting the IsKeyless property to false before defining the key, e.g.
e.HasNoKey(); // generated by scaffolding
e.Metadata.IsKeyless = false; // <--
e.HasKey(o => o.OrderDetailId); // now this works
I'm manually moving from .NET Framework to .NET Core and working on the EF and DTOs.
I'm not enjoying this - I have read indexes are not supported via annotation and as such, am currently mixing fluent api and annotations which is a code smell. However, it appears as if I must proceed with this combination.
My question is if I can achieve this only with fluent api. My table has both a primary key and a unique constraints.
My object looks like
class Person
{
public int Id {get;set;}
public string UniqueToken {get;set;}
}
However, I am unable to add the following
modelBuilder.Entity<Person>()
.HasKey(a => a.Id)
.HasIndex(a => a.UniqueToken).IsUnique(); //this is what I would like to add but I can't.
I've attempted something which feels like a hacky work around
modelBuilder.Entity<Person>()
.HasKey(a => a.Id);
modelBuilder.Entity<Person>()
.HasIndex(a => a.UniqueToken).IsUnique();
Again, adding this entry twice seems a little bleugh… Fluent appears to want to simply chain the methods.
I have read on this, but I'm getting very lost. Is it possible to add both the primary key and unique constraint ?
Better you separate your entity configurations from OnModelCreating method with IEntityTypeConfiguration interface as follows:
public class PersonConfiguration : IEntityTypeConfiguration<Person>
{
public void Configure(EntityTypeBuilder<Person> builder)
{
builder.HasKey(a => a.Id);
builder.HasIndex(a => a.UniqueToken).IsUnique();
}
}
Now you can add configuration for all of your Db entities like PersonConfiguration and register all of them at once as follows:
protected override void OnModelCreating(ModelBuilder modelBuilder)
{
base.OnModelCreating(modelBuilder);
modelBuilder.ApplyConfigurationsFromAssembly(typeof(PersonConfiguration).Assembly);
}
This will give you more separation of concern and readability!
I need to load records from an existing database into a new database using a LINQ query using LINQPad. If the record exists in the database, then update the name. Otherwise, insert the record. Currently, the new DB is empty. This script will run periodically, so I have to check for updates. I have code that loads the existing and new records into two List for comparison. I loop through the list:
foreach (Coaster oldCoaster in listOfOldCoasters) {
var coaster = listOfNewCoasters.Where(c => c.coasterId == oldCoaster.coasterId).FirstOrDefault();
if (coaster != null) {
coaster.Name = oldCoaster.Name;
} else {
coaster = new Models.Coaster();
coaster.CoasterId = oldCoaster.coasterId;
coaster.Name = oldCoaster.Name;
//newCoasterDbContext.Coasters.Attach(coaster);
newCoasterDbContext.Coasters.Add(coaster);
}
}
newCoasterDbContext.SaveChanges();
When I run the code using the "Add" method, I receive the exception "OriginalValues cannot be used for entities in the Added state." Digging deeper, I see this message:
Cannot insert the value NULL into column 'CoasterId', table 'Coaster'; column does not allow nulls.
I am setting the primary key in question, so I must be missing something about EF as to why this fails. If I uncomment the "Attach" statement and try that instead of "Add", then the script runs, but nothing gets inserted into the database.
My Coaster class:
public class Coaster
{
public System.Guid CoasterId { get; set; }
public string Name { get; set; }
public Coaster()
{
CoasterId = System.Guid.NewGuid();
}
}
Based on other posts, I have tried adding attributes to the PK:
[Key]
[DatabaseGenerated(DatabaseGeneratedOption.None)]
And I have tried modifying the OnModelCreating method:
HasKey(x => x.CoasterId).Property(x => x.CoasterId).HasDatabaseGeneratedOption(System.ComponentModel.DataAnnotations.Schema.DatabaseGeneratedOption.None);
These haven't worked for me. What have I missed that is causing Entity Framework to ignore the GUID that I have attached to the primary key field?
You can't fix stupid, but you can fix stupid mistakes. So as I mentioned, in the OnModelCreating method, I had this line:
HasKey(x => x.CoasterId).Property(x => x.CoasterId).HasDatabaseGeneratedOption(System.ComponentModel.DataAnnotations.Schema.DatabaseGeneratedOption.None);
I thought it didn't work, but that was only because a bit later in the method, there was a line that overwrote change:
Property(x => x.FpoDistrictId).HasColumnName(#"FPODistrictId").HasColumnType("uniqueidentifier").IsRequired().HasDatabaseGeneratedOption(System.ComponentModel.DataAnnotations.Schema.DatabaseGeneratedOption.Identity);
At least I figured it out...
I'm having problems with duplicate data during migration with Code First.
A new foreign key record is duplicated each time the migration creates the master record.
The schemas in the database are being created correctly. Namely the Primary Keys and Foreign Key values (the latter being automatically generated)
Can someone please advise thanks about how I detach the foreign key record during migration to prevent it recreating the record or any configuration I need to implement? I've tried updating the state of the foreign key obects before inserting master data. to both modified and detached.
For example I see multi records for the same priority where there should only be 3.
I'm using Entity Framework 6.0.
public class VeloPointDbConfiguration : DbMigrationsConfiguration<VeloPointDbContext>
{
public VeloPointDbConfiguration()
{
AutomaticMigrationsEnabled = true;
AutomaticMigrationDataLossAllowed = true;
}
protected override void Seed(VeloPointDbContext context)
{
context.TaskPriorities.AddOrUpdate(EventTaskPriority.Migrations.All());
context.TaskStatuses.AddOrUpdate(TaskStatus.Migrations.All());
EventOrganiserTask.Migrations.All().Select(x => context.Entry(x.Priority).State == EntityState.Modified);
EventOrganiserTask.Migrations.All().Select(x => context.Entry(x.TaskStatus).State == EntityState.Modified);
context.Tasks.AddOrUpdate(EventOrganiserTask.Migrations.All());
}
}
The following examples of the instances i'm using for the data.
I create the following methods for the foreign key objects
public static EventTaskType[] All()
{
return new[]
{
GetDeadline(),
GetEmail(),
GetTelephone(),
GetAppointment(),
GetSnailMail(),
};
}
internal static EventTaskType GetDeadline()
{
return new EventTaskType("09974722-D03E-4CA3-BF3A-0AF7F6CA1B67", 1, "Deadline")
{
Icon = ""
}
}
I call the following methods the create the master data.
public static EventOrganiserTask[] All()
{
return new EventOrganiserTask[]
{
GetBookHQ(1, new DateTime(Event.Migrations.EventDate.Year - 1, 10, 1)),
GetFindSponsor(2, new DateTime(Event.Migrations.EventDate.Year - 1, 10, 1)),
GetRegisterEvent(3, new DateTime(Event.Migrations.EventDate.Year - 1, 10, 1)),
GetBookFirstAid(4, Event.Migrations.EventDate.AddMonths(-6))
};
}
NOTE: When creating the master record, I call the method in the foreign key classes each time - which is the crux of the problem where I need to instruct the migration to detach this item.
public static EventOrganiserTask GetRegisterEvent(int id, DateTime date)
{
return new EventOrganiserTask
{
id = id,
Title = "Register event",
Summary = "Register the road race with the region",
DueDate = date,
Priority = EventTaskPriority.Migrations.GetHighPriority(),
Person = Person.Migrations.GetRaceOrganiser(1),
TaskType = EventTaskType.Migrations.GetDefault(),
TaskStatus = TaskStatus.Migrations.GetDefault(),
};
}
NOTE: When I do make changes to the data from the application, the foreign keys are not being updated. This must be related and indicates my entities are not configured correctly.
LATEST:
I'm still tearing my hair out. I've investigated this further and read about the migrations being multi threaded (It was another thread on stackoverflow but I can't find it again). Indeed running the Seed method I supposed is what it says on the tin and is purley for seeding data, so the data is only being added (regardless of AddOrUpdate - what's that all about then) So I've looked at the behaviour regarding the records being created. First of all I called context.SaveChanges() after creating the look up tables. At this point it doesn't created any duplicates as the items are only referenced once. I then let the seed method run the master data - but argggh - I see duplicates (when the instances are called on the master data). But this did flag something up with regard to the order in which it creates the records.
My next step was to create two migrations, but without any success.
I'm hoping somebody picks up this thread soon. I'm tearing my hair out.
Ok so i've finally found my answer. It was clever enough to create the foreign key relationships from the model, but i needed to be explicitly set the foreign key id field. I chose the Fluent API to explicitly set the relationships and I set the value of the id field in the mapping of the object.
modelBuilder.Entity<Task>()
.HasRequired(x => x.Priority)
.WithMany(x => x.Tasks)
.HasForeignKey(x => x.Priority_id);
Here it is in the seed method
public class VeloPointDbConfiguration : DbMigrationsConfiguration<VeloPointDbContext>
{
public VeloPointDbConfiguration()
{
AutomaticMigrationsEnabled = true;
AutomaticMigrationDataLossAllowed = true;
}
protected override void Seed(VeloPointDbContext context)
{
context.TaskPriorities.AddOrUpdate(EventTaskPriority.Migrations.All());
context.TaskStatuses.AddOrUpdate(TaskStatus.Migrations.All());
EventOrganiserTask.Migrations.All().Select(x => context.Entry(x.Priority).State == EntityState.Modified);
EventOrganiserTask.Migrations.All().Select(x => context.Entry(x.TaskStatus).State == EntityState.Modified);
context.Tasks.AddOrUpdate(EventOrganiserTask.Migrations.All());
// Foreign Key relationships
modelBuilder.Entity<EventOrganiserTask>()
.HasRequired(x => x.TaskStatus)
.WithMany(x => x.Tasks)
.HasForeignKey(x => x.TaskStatus_id);
modelBuilder.Entity<Task>()
.HasRequired(x => x.TaskType)
.WithMany(x => x.Tasks)
.HasForeignKey(x => x.TaskType_id);
modelBuilder.Entity<Task>()
.HasRequired(x => x.Priority)
.WithMany(x => x.Tasks)
.HasForeignKey(x => x.Priority_id);
}
}
How do I setup the mapping for these table relationships using code first and the fluent api?
I have a poco for all three entities created but most people don't have a poco for the "intermediate" table (PlanUser) so I couldn't find a good example to go off of. I need to tell EF to map the PlanUser.PlanCustomerId column to Plan.CustomerId but it either doesn't return the right results OR as in the current setup the mapping for Plan throws the error:
"The specified association foreign key columns 'CustomerId' are
invalid. The number of columns specified must match the number of
primary key columns."
public class PlanUserMap : EntityTypeConfiguration<PlanUser>
{
public PlanUserMap()
{
this.ToTable("PlanUser");
this.HasKey(c => new { c.CustomerId, c.PlanCustomerId });
this.HasRequired(c => c.Plan).WithMany().HasForeignKey(x => x.CustomerId).WillCascadeOnDelete(false);
this.HasRequired(c => c.Customer).WithMany().HasForeignKey(x => x.CustomerId).WillCascadeOnDelete(false);
}
}
public class PlanMap : EntityTypeConfiguration<Plan>
{
public PlanMap()
{
this.ToTable("Plan");
this.HasKey(c => c.CustomerId);
// the below line returns only 1 row for any given customer even if there are multiple PlanUser rows for that customer
//this.HasMany(c => c.PlanUsers).WithRequired().HasForeignKey(c => c.PlanCustomerId);
// this throws an error
this.HasMany(c => c.PlanUsers).WithMany().Map(m => m.MapLeftKey("PlanCustomerId").MapRightKey("CustomerId"));
}
}
public partial class CustomerMap : EntityTypeConfiguration<Customer>
{
public CustomerMap()
{
this.ToTable("Customer");
this.HasKey(c => c.Id);
this.HasMany(c => c.PlanUsers).WithRequired().HasForeignKey(c => c.CustomerId);
}
}
#Slauma, sql profiler shows these queries being executed. The second one should include customer id 43 in addition to customer id 1 but it does not. I don't know why its not retrieving that second row.
exec sp_executesql N'SELECT
[Extent1].[CustomerId] AS [CustomerId],
[Extent1].[PlanCustomerId] AS [PlanCustomerId],
[Extent1].[CreatedOnUtc] AS [CreatedOnUtc],
[Extent1].[IsSelected] AS [IsSelected],
[Extent1].[IsDeleted] AS [IsDeleted],
[Extent1].[AccessRights] AS [AccessRights]
FROM [dbo].[PlanUser] AS [Extent1]
WHERE [Extent1].[CustomerId] = #EntityKeyValue1',N'#EntityKeyValue1 int',#EntityKeyValue1=43
exec sp_executesql N'SELECT
[Extent1].[CustomerId] AS [CustomerId],
[Extent1].[Name] AS [Name],
[Extent1].[PlanTypeId] AS [PlanTypeId],
[Extent1].[OrderId] AS [OrderId],
[Extent1].[CreatedOnUtc] AS [CreatedOnUtc],
[Extent1].[IsActive] AS [IsActive]
FROM [dbo].[Plan] AS [Extent1]
WHERE [Extent1].[CustomerId] = #EntityKeyValue1',N'#EntityKeyValue1 int',#EntityKeyValue1=1
Here is the C# code that causes the queries to execute:
public List<Plan> GetPlans()
{
List<Plan> plans = new List<Plan>();
// add each plan they have access rights to to the list
foreach (var accessiblePlan in Customer.PlanUsers)
{
plans.Add(accessiblePlan.Plan);
}
return plans;
}
I think what you need is actually simpler than the mapping you tried:
public class PlanUserMap : EntityTypeConfiguration<PlanUser>
{
public PlanUserMap()
{
this.ToTable("PlanUser");
this.HasKey(pu => new { pu.CustomerId, pu.PlanCustomerId });
this.HasRequired(pu => pu.Customer)
.WithMany(c => c.PlanUsers)
.HasForeignKey(pu => pu.CustomerId)
.WillCascadeOnDelete(false);
this.HasRequired(pu => pu.Plan)
.WithMany(p => p.PlanUsers)
.HasForeignKey(pu => pu.PlanCustomerId)
.WillCascadeOnDelete(false);
}
}
public class PlanMap : EntityTypeConfiguration<Plan>
{
public PlanMap()
{
this.ToTable("Plan");
this.HasKey(p => p.CustomerId);
}
}
public partial class CustomerMap : EntityTypeConfiguration<Customer>
{
public CustomerMap()
{
this.ToTable("Customer");
this.HasKey(c => c.Id);
}
}
I am not sure why you disable cascading delete. I probably wouldn't do this (i.e. I would remove the WillCascadeOnDelete(false) for both relationships) because the association entity PlanUser is kind of dependent of the other two entities.
Here are a bit more details about this kind of model (sometimes called "many-to-many relationship with payload"): Create code first, many to many, with additional fields in association table
Ok, I figured it out. Thanks to #Slauma for the help! We are using the repository pattern with IOC(autofac). I shouldn't be calling Customer.PlanUsers in my repository service in the first place. Somewhere along the line we started resolving the Customer entity using IOC in the repository service to get at certain customer properties quickly and put it in a read-only variable. Inevitably this hosed us because instead of querying the entity repositories themselves we started with the Customer entity expecting it to be filled with everything we needed and its not pre-filled.
public List<Plan> GetPlans()
{
List<Plan> plans = new List<Plan>();
var list = _planUserRepository.Table.Where(a => a.CustomerId == Customer.Id).ToList();
foreach (var planUser in list)
{
plans.Add(planUser.Plan);
}
}